Silicon Valley has found a new way to sweeten job offers: AI tokens. Instead of traditional signing bonuses, tech companies are quietly offering engineers massive budgets of computational credits — the currency that powers ChatGPT, Claude, and other AI tools. Nvidia CEO Jensen Huang suggested this week that top engineers should receive roughly half their base salary again in tokens, potentially $250,000 worth annually.
The shift reflects how dramatically AI usage has changed among software engineers. Where someone writing an essay might burn through 10,000 tokens in an afternoon, engineers running swarms of automated agents can consume millions in a single day — all happening in the background while they sleep.
The trend gained momentum after the January release of OpenClaw, an open-source AI assistant designed to run continuously. Unlike traditional AI tools that respond to specific prompts, OpenClaw spawns sub-agents and works through task lists autonomously — dramatically increasing token consumption in the process.
Venture capitalist Tomasz Tunguz of Theory Ventures was among the first to identify this shift in February, noting that startups were adding "inference costs as a fourth component to engineering compensation." Using data from Levels.fyi, he calculated that a top-quartile software engineer earning $375,000 might receive an additional $100,000 in tokens — meaning roughly one dollar in five becomes compute power.
But the apparent generosity comes with strings attached. Companies funding substantial token budgets naturally expect proportional productivity gains — effectively requiring engineers to perform at twice their usual rate if they're receiving compute worth half their salary.
The financial implications get murkier when token spending approaches or exceeds employee salaries. If AI agents are doing much of the work, companies may start questioning how many human coordinators they actually need.
Financial services CFO Jamaal Glenn, a former VC and Stanford MBA, points to another concern: tokens as compensation inflation without actual value transfer. Unlike salary or equity, token budgets don't vest, appreciate, or carry forward to future negotiations.
"What may seem like a perk can be a clever way for companies to inflate the apparent value of a compensation package without increasing cash or equity — the things that actually compound for an employee over time," Glenn notes.
If companies successfully normalize tokens as standard compensation, they could potentially keep cash compensation flat while pointing to growing compute allowances as evidence of investment in their workforce — a structure that clearly benefits employers more than employees.
The practice raises fundamental questions about the nature of compensation in an AI-driven economy. Traditional benefits like health insurance and retirement matching represent genuine employee value. Token budgets, by contrast, primarily serve to subsidize the AI infrastructure that makes engineers more productive for their employers.
Whether this represents genuine added value or clever accounting depends largely on questions most engineers don't yet have enough information to answer — including how token economics will evolve as AI capabilities advance and potentially reduce the need for human oversight altogether.